If your code is local, you can use COPY to copy it into the image, and then RUN pip install locally. All this you can do in a patch for the Dockerfile, so you don’t have to edit it manually.
Not really. That means my new_api folder must live in the same context, which it does not. See Build context | Docker Docs
So it means I must copy my new_api into the folder where the dockerfile is.
Plus, need to make sure the copied folder into the stage gets copied over into the images. See here